Optimization Of Rail Profile On Curve Based On The Competitive Relationship Between Fatigue And Wear

With the increasing axle load of heavy-haul railway and the increasing density of cargo transportation,the rail damage is becoming more and more serious,especially the problem of rolling contact fatigue?RCF?and wear of rail is particularly prominent.According to the competitive failure mechanism of rail wear and rolling contact fatigue in heavy-haul railway,a numerical optimization method of rail profile using different optimization strategies for inter and outer rails is proposed to optimize rail profile,improve wheel and rail relationship,reduce rail fatigue and wear damage,and prolong rail service life.The main research work of this thesis is as follows:A prediction model of rail wear and a prediction model of fatigue crack propagation are established.Based on the theory of vehicle-track coupling dynamics,the simulation prediction model of rail wear is established on the basis of Archard wear model and multi-point Kik-Piotrowski?KP?Contact theory.Using the energy index?EI?method and Miner linear fatigue damage accumulation theory,the prediction model of rail fatigue crack propagation is established,and the validity of the prediction model of rail wear and fatigue crack propagation is verified by using the field measured rail damage data.This thesis analyzes the main factors affecting rail wear and rolling contact fatigue,and holds that it is considered that the influence of truck marshalling model,wheel profile,track irregularities and running speed should be considered when predicting rail wear and fatigue.The design parameters of rail profile optimization are determined.The prediction model of rail wear and the prediction model of fatigue crack propagation under various stochastic factors are established by using stochastic sample method,and the development law of 75kg/m standard profile rail wear and fatigue crack is studied under stochastic factors,and the design period and inter and outer rail optimization area of rail profile optimization are determined.The nonlinear mathematical optimization model of rail profile is established by using approximate model method.Using Latin hypercube sampling?LHS?method to extract the rail profile sample,taking the rail profile arc parameter as the independent variable,taking the rail fatigue and wear balance as the optimization goal in the design cycle,using two approximate model methods of BP neural network and Gaussian Radial Basis Function?RBF?,the optimization model of rail profile is established,and the optimization model established by the two methods has high precision.The prediction error is less than 5%.Two optimal inter and outer rails combinations of profiles GS₁ and BP₁ are obtained by using genetic algorithm?GA?to solve the optimization model of two kinds of rail profiles.During the design cycle,the wear rate and crack propagation rate of GS₁ combined profile and BP₁ combined profiles are always dynamically balanced,and the surface of rail does not appear fatigue damage.Compared with the 75kg/m standard profile,the optimal profile wheel and rail contact point moves to the non-working side,the rail wear is more uniform,the rail contact spot area increases,the contact stress of the wheel and rail decreases 1¹.4 times,and the service life of the rail is extended by 10%and 19.4%respectively during the design cycle.